Fuzzy Fusion of Change Vector Analysis and Spectral Angle Mapper for Hyperspectral Change Detection

Change Vector Analysis (CVA) is one of the most widely used approaches for change detection in multispectral and hyperspectral images. Although, in CVA, the spectral change vector (CV) comprises the angle as well as the magnitude of the change, typically only the magnitude measure is used as change criterion. On the other hand, the spectral angle mapper (SAM) uses only the angle measure as criterion for change detection. It is envisaged that combining the angle and magnitude for change detection (i.e. combining SAM and magnitude CVA) can improve the change detection performance, yet only a limited number of approaches have been proposed in the literature so far. This paper presents a novel fuzzy inference combination strategy that combines the angle and magnitude distances, referred to as Fuzzy CVA (FuzCVA), and is shown that the proposed approach can provide improved change detection performance by effectively combining magnitude and angle measures.